460 ANNUAL REPORT SMITHSONIAN INSTITUTION, 195 5 



studying variations such as those of the markings on Mars, that the 

 rings of trees might directly or indirectly record year-by-year changes 

 in the sun. 



Julius C. Kapteyn, about 1880, examined oak sections from western 

 Germany and Holland and derived a 240-year ring chronology. The 

 later decades of this history corresponded well with rainfall data, 

 and a strong, but quite unexplainable, cyclic variation of 12.4 years 

 was present throughout. Fortunately for astronomy, Kapteyn evi- 

 dently felt this work to be strictly extracurricular, for he carried it no 

 further. But his single paper on the subject, a published lecture in 

 Pasadena in 1908, is a delight to read for its simple presentation of 

 essentials and its humility. 



In contrast to this somewhat abortive effort, the program of re- 

 search initiated by A. E. Douglass at Flagstaff, Ariz., in 1904, has 

 been carried on for almost half a century at the University of Arizona, 

 at Tucson, and has led to important developments in quite unexpected 

 directions. The most spectacular development was a method which 

 made it possible to precisely date many ruins and thus provide a time 

 scale for the pre-Spanish cultures of tlie Southwest. This method is 

 based on detailed matching or cross-dating of the patterns in tree 

 rings — an application of the operation of forecast-and-verification, 

 which is such an integral part of the scientific method. It proved 

 to have far-reaching implications in climatic studies as well, for it 

 was the essential key to the development of highly significant tree- 

 ring histories of rainfall and other climatic variables. 



We thus see that modern techniques in dendrochronology find their 

 principal application in two fields of research: (a) dendroclimatol- 

 ogy, that is, historical climatology based on fluctuations in ring 

 growth, and (h) dendroarcheology, the dating of prehistoric struc- 

 tures and activities by the precise dating of ancient wood. Before we 

 consider the contributions in these two fields, however, it may be 

 illuminating to directly examine some tree rings of the sensitive type 

 useful in dendrochronology. 



The photographed outer rings in radial increment cores from the 

 lower stem of selected living Douglas-fir trees are shown in plate 1. 

 These represent the ring chronologies in recent decades in three locali- 

 ties of the Colorado River basin — the isolated and arid Nine Mile 

 Canyon of northeastern Utah, the archeologically rich Mesa Verde of 

 southwestern Colorado, and the Tucson area of southern Arizona — a 

 range of over 500 miles north-south. 



In the outermost ring at the extreme right in the upper photograph 

 of plate 1, the band of light-colored early wood must have been laid 

 down largely during June and early July, as is characteristic of moun- 

 tain conifers near the lower forest border throughout the western 

 United States. The growth of dark latewood cells, which complete 



